Infrared Imaging with John Charles

In this Technical Forum, we'll discuss some of the techniques, equipment, and opportunities for infrared imaging (also referred to as "thermography") in the paper mill, and some of the pitfalls to avoid along the way. Before we start, let's take an overview of the field.

There is a great deal written about infrared (IR) imaging these days. We see news stories in which infrared cameras are used to find lost hikers and bad guys evading the police at night. We watched an incredible display of the use of infrared imaging during the Persian Gulf War. The recent fires in Florida were attacked by fire fighters using infrared imaging to locate "hot spots" that could flare up again later. We even see TV commercials that use infrared imaging to show the heat coming from arthritic joints and the impact of spreading this or that kind of magic "goo" on them. Such casual use of infrared images creates the impression that IR technology and its use are simple, every-day events that require little or no interpretation to be understood. Sometimes that's true. More often it's not.

The industrial uses of infrared imaging are much more serious business, with some industries depending heavily on the technology for safety, productivity, and diagnostics. Industrial infrared imaging has come a long way since the early 70's when little of the technology was unclassified and less was commercially available. The medical professions were quick to embrace infrared imaging as a diagnostic tool. My own early exposure to infrared imaging came at the University of Oklahoma while pursuing my doctorate in Mechanical Engineering in 1972. We acquired a Texas Instruments system that produced very detailed black-and-white thermograms over a limited temperature range. It was the size of a small refrigerator and needed liquid nitrogen for cooling its detector. Despite these logistical difficulties we were actively engaged in research with the OU Medical School studying cancer and circulatory disorders.

Some of the other work we pursued was more industrial in nature. Flaw detection in composite materials (such as graphite/epoxy and fiberglass) is one example. By heating the material, one can find internal delaminations before they produce catastrophic failures. The aerospace industry still uses a similar technique to proof test composite wing panels for fighter aircraft. Other uses of infrared imaging took advantage of the heat inherent in some processes or materials to provide diagnostic information. A small industry of "inspectors" sprang up using IR imagers to look for heat "leaks" from houses and the heat generated by poor electrical connections in power stations. Gradually the equipment needed for infrared imaging became smaller, more sophisticated and yet more user-friendly to meet the growing market for infrared inspection. Today there are infrared imagers on the market which require no external cooling and are the size of a small camcorder. They produce great results and interface nicely with PC's for storage of data and reporting results.

So where does the paper industry fit in to the infrared imaging world? If you've ever visited or worked in a paper mill, you know there's lots of heat to look at out there! And just as there are many heat sources, there are many opportunities to use IR imaging in the paper mill to solve problems and provide reliability and safety. Many mills contract with vendors to provide electrical distribution panel surveys periodically. Some clothing manufacturers are now providing IR imaging as a service to customers. The potential is only limited by our imagination and ingenuity. With that in mind, let's proceed with some frequently asked questions about infrared imaging.

What does "infrared" mean?

"Infrared" refers to the wavelength of the energy emitted by a body due to it's temperature, which lies just below the visible red light wavelength in the electromagnetic spectrum. On the other end of the electromagnetic spectrum are radio waves. We can see only a small part in the middle of the spectrum with the naked eye (i.e., the "visible" portion), but sensors have been developed that allow us to detect the other wavelengths with instruments.

How is infrared energy produced?

IR energy is emitted from everything that has a temperature above absolute zero (which means just about everything!). The energy emitted is proportional to the temperature of the object raised to the fourth power.

How is IR energy transmitted?

Unlike conductive heat transfer (which takes place through solid materials) or convective heat transfer (which uses fluid to move the heat), radiative heat transfer takes place through transparent media in a manner analagous to light transmission. The best medium for transmitting IR energy is a vacuum. In the presence of air, some of the energy is absorbed by the air itself and some is scattered. High quality equipment used for IR imaging takes this effect into account when processing radiation.

How is IR energy detected?

Infrared detectors are solid-state devices which produce a signal proportional to the total radiation received. Imager detectors are cooled to have the sensitivity and fast response time necessary for viewing a moving field of temperatures. Early cooling was via liquid nitrogen and argon gas. Today, imagers use Sterling coolers or a thermo-electric effect to lower the detector temperature. The small amounts of radiation received are amplified by signal processors to produce the image itself, which is a false-color replication of the temperature field.

Can't I get the same results using a spot radiometer, or "heat gun"?

No, although spot radiometers work in a similar manner, they are slow to respond to changes and read from only one spot in the field of view. Even with an array of spot radiometers you could only get the temperature distribution of a very slowly changing temperature field. However, spot radiometers are good for determining the average temperature of moving objects and have found widespread use in paper mills. But don't confuse infrared imagers with spot radiometers - they are completely different animals.

How fast do IR imagers respond compared to spot radiometers?

To answer this question completely we will have to discuss the two modes of operation for infrared imaging - "full field" and "line scan". For now we'll think just about the TV-like infrared images of the objects around us. Most imagers respond to changes fast enough to keep the picture generated "real-time". This means a response of at least 25 Hertz (Hz), or 25 frames per second. Most spot radiometers respond with a time constant of about 1/3 of a second, or 3 Hz. Much higher rates of data acquisition are available in line-scan capable scanners, some up to 3500 Hz.

How is the energy received by the detector converted to temperatures?

This question gets to a fundamental point about infrared temperature detection via any means. That is, what is actually detected is incoming radiation, or the "radiance" of the object, not temperature. To convert the radiance into temperatures you need to know several things about the object you're looking at. These include details about the response of the detector - what signal is produced per radiance unit received at what wavelength, etc. You also have to know about the air the radiance is transmitted through - the atmospheric temperature and humidity are important factors in determining how much energy radiated from the object is absorbed by the air. You also have to know the distance from the object for atmospheric absorption calculations. Finally, you need to know the object's surface emissivity. As you can see, determining the precise temperature of an object can be difficult, especially when some of the parameters in the equation have to be estimated.

What exactly is emissivity?

Emissivity is a unitless factor between 0 and 1 which describes the fraction of available thermal radiation (due to an object's temperature) which is emitted from an object. Items which emit all the radiation possible have an emissivity of 1 and are called "black bodies". They also have the property of absorbing all the radiation that strikes them, so there is no radiation reflected from the surface. Objects with an emissivity of zero theoretically emit nothing due to their own temperature (which makes them poor candidates for IR imaging!), but reflect everything that they receive. Emissivity is largely a property of the surface of an object - rough surfaces tend to have high emissivity, smooth ones lower values. Note that emissivity is not a "color" property of the surface. For example, snow is a black body with an emissivity of 1.

How is emissivity determined?

The best way to determine emissivity is to "measure" it by comparing an infrared image with a contact pyrometer which physically measures the surface temperature of the object. Any differences must be accounted for by emissivity (and transmission losses). Making this measurement is not always easy to do, but happily there is data available for many common surfaces, including paper and metals common in the mill.

A note of caution: Sometimes the emissivity of an object changes from point to point over a surface due. This may be the result of coatings or dirt on the surface. When this happens, infrared images can seem to display a wide range of temperatures when in fact the temperature of the object is uniform. Some systems have area compensation for emissivity changes, but most don't. Always evaluate the surface being scanned for uniformity before accepting the display from an infrared imager!

What if I don't have a good emissivity value?

Much of the time, what matters most in infrared imaging is the pattern of temperatures rather than the temperatures themselves. If you don't get fooled by the emissivity variations mentioned above, you can still get good information from the images. For example, hot spots at motor lead connections could mean a bad connection whether the absolute temperature is 100 degrees or 150 degrees. Cold regions in a reel of paper are generally wetter than surrounding regions regardless of their absolute temperature. Certainly, knowing accurate temperatures gives more information that can be used to better diagnose a problem, but small errors in emissivity don't negate the value of infrared imaging in the mill.

What are some of the paper mill uses of infrared imaging?

There are quite a few places where infrared imaging can be very useful in the paper mill. Above we mentioned the inspection of motor connections and control centers. Power distribution panels are good candidates for thermal inspection. Other areas that are suited to IR inspection are tank and duct insulation, bearing housings, and steam traps. There are numerous places in the pulp mill where IR imaging is also appropriate. Many of these services are available from consultants in most areas.

Infrared imaging on the paper machine is an area we specialize in at PROdry Technology, Inc. The applications break down into two groups according to the technique used to address them. The first uses "full-frame" imaging to visualize temperatures in a TV-like view. The second group uses a technique called "line scanning" to get high-speed images of moving objects.

What is "line scan" infrared imaging?

When you look at a full-frame infrared image, you see a picture of the temperature distribution in the field of view, much like a TV picture. If the object is moving, you see an average temperature at each point for the data acquisition time (1/25 second). The object can appear blurred due to its own motion. A case in point is a Yankee dryer, which may rotate at 1 revolution each 2/3 of a second.

In contrast, line scanning continuously looks at a single line in the field of view (which may have 150 to 300 lines total) as an object moves by. Since only one line is scanned, the data acquisition rate goes way up. Our own IR system scans a line at close to 3500 Hz, which is about 140 times faster than full-frame imaging. This high-speed scanning effectively freezes the object passing the scan line. Continuously scanning the same line produces an "extruded" image of the temperature on the surface. As an example, for the highest speed tissue machines today, the result would be an image with 0.34 inches of the dryer seen by each line, and the full extruded image would be made up of about 2000 lines! With this high spatial resolution, line scanning offers unique opportunities for diagnosing problems in the mill when they involve moving objects

Of course, the systems capable of line scanning are more expensive to purchase and operate than standard imagers. But look at the examples under the Infrared Imaging page of this site for some good line-scanned IR images and you'll see why line scanning is the preferred means of diagnosing many paper machine problems.

If that's true, is full frame imaging ever desirable?

Sure, especially if the object you're looking at isn't moving. It would make no sense to line scan a duct to look for insulation problems, for example. Full frame imaging is useful in addressing problems on the machine that are steady in the machine direction: wet streaks from showering, basis weight streaks, general profile problems from pressing, and so on. Clothing manufacturers commonly use full frame imaging to look at felts to see if there are thermal streaks in them. Looking at the reel or web along the machine can give insight to moisture streaks with higher sensitivity than we get from dry-end scanners.

When is line scanned imaging desirable?

The simple answer is "whenever there can be MD temperature variations." As a rule, full-frame IR imaging can be used effectively to view temperature patterns that vary in the cross-machine direction, but do not change in the machine direction. For example, viewing a felt can give good information about showering problems that are steady CD. But if MD problems such as barring and contamination exist in the felt, full-frame imaging likely won't pick them up. But line scanning allows us to see the entire felt temperature distribution in both CD and MD.

One of the best applications of line scanning is looking at Yankee dryers. Images captured in line scan mode yield much information as to the performance of the internal condensate removal system. Furthermore, line scanned images synchronized to Yankee rotation make it possible to identify which specific condensate header is creating a problem. Individual plugged or broken dipper tubes can be pinpointed and slated for repair at the next outage.

You can also get good results scanning the sheet itself. In some cases this is necessitated by the incredible amount of equipment we continue to "hang" around the Yankee dryer. It is sometimes impossible to get any view of the dryer itself, however limited. Fortunately, the sheet carries the thermal impression of the yankee dryer very well, and factoring crepe and draw into the image can result in accurate location of problems on the dryer. The sheet can also give up information about itself when viewed in line scan. For example, cockle in fine paper shows up prominently when viewed with line scanners. Processing the images with Fourier transforms can give frequency data which can point to rotational elements responsible for the problem.